Psychological Review © 2011 American Psychological Association

2011, Vol. 118, No. 3, 482– 495 0033-295X/11/$12.00 DOI: 10.1037/a0023477

Relational Regulation Theory: A New Approach to Explain the Link

Between Perceived Social Support and Mental Health

Brian Lakey Edward Orehek

Grand Valley State University University of Groningen

Perceived support is consistently linked to good mental health, which is typically explained as resulting
from objectively supportive actions that buffer stress. Yet this explanation has difficulty accounting for
the often-observed main effects between support and mental health. Relational regulation theory (RRT)
hypothesizes that main effects occur when people regulate their affect, thought, and action through
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

ordinary yet affectively consequential conversations and shared activities, rather than through conver-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

sations about how to cope with stress. This regulation is primarily relational in that the types of people
and social interactions that regulate recipients are mostly a matter of personal taste. RRT operationally
defines relationships quantitatively, permitting the clean distinction between relationships and recipient
personality. RRT makes a number of new predictions about social support, including new approaches to
intervention.

Keywords: social support, perceived support, affect, emotion regulation, SRM

The social support literature is vast, as reflected in over 33,000 the link between life stress and poor mental health is stronger for
entries in PsycINFO and over 45,000 in Medline. It is scattered people with low social support than for people with high social
across a wide range of disciplines, including psychology, psychi- support. A key idea is that in the absence of stress, social support
atry, medicine, nursing, social work, sociology, and communica- is not linked to mental health (see Figure 2 in Cohen & Wills,
tions. Social support is an important construct because it has been 1985). Main effects occur when people with high social support
linked to a wide range of health outcomes, including both mental have better mental health than those with low social support,
(Barrera, 1986; Cohen & Wills, 1985) and physical health regardless of stress levels.
(Uchino, 2004, 2009). Perceptions that family and friends would Stress buffering theory has been thoroughly developed and has
provide effective help during times of stress (i.e., perceived sup- dominated social support research. Nearly all research on social
port) have been consistently linked to good mental health, includ- support is guided by the assumption that social support’s link to
ing low rates of major depression (Lakey & Cronin, 2008), few mental health reflects stress buffering. Stress buffering theory is
posttraumatic stress disorder symptoms (Brewin, Andrews, & Val- not a single author’s work, and many scholars have contributed to
entine, 2000), low levels of nonspecific psychological distress its development. Some of the clearest, and most widely cited
(Barrera, 1986; Cohen & Wills, 1985; Procidano, 1992), and low statements are those of Barrera (1986), Cohen and Wills (1985),
negative affect, as well as high positive affect (Finch, Okun, Pool, Cutrona and Russell (1990), and Thoits (1986). Stress buffering
& Ruehlman, 1999). theory is an extension of the general stress and coping theory of
Lazarus (1966) and Lazarus and Folkman (1984). As applied to
Research and Theory on Stress Buffering mental health, the theory can be summarized succinctly by five
Cohen and Wills (1985) distinguished between social support’s hypotheses: (a) Life events are stressful to the extent that people
stress buffering and main effects, and this has played a founda- perceive the events as threats (i.e., primary appraisal) and perceive
tional role in shaping research and theory. Stress buffering occurs themselves as lacking adequate responses (i.e., secondary ap-
when social support protects (i.e., buffers) people from the bad praisal). (b) Events increase risk for poor mental health depending
effects of stress. Evidence for stress buffering is indicated when on people’s coping. Coping involves a wide range of deliberate
thought and action, including problem solving, reappraisal, avoid-
ance, and support seeking. (c) Social support is a relatively stable
This article was published Online First May 2, 2011. resource that buffers stress, primarily by influencing appraisal and
Brian Lakey, Psychology Department, Grand Valley State University; coping. (d) Social support includes what friends and family say
Edward Orehek, Department of Psychology, University of Groningen, and do regarding stressful events (i.e., enacted support), as well as
Groningen, the Netherlands. recipients’ perceptions that quality enacted support is available
Lawrence H. Cohen, Robert Hendersen, and Edward Lemay provided (i.e., perceived support). Perceived support is based primarily on
valuable comments on an earlier draft of this manuscript.
one’s history of receiving effective enacted support. (e) Social
Correspondence concerning this article should be addressed to Brian
Lakey, Psychology Department, Grand Valley State University, One Cam- support is effective in buffering stress when the support specifi-
pus Drive, Allendale, MI 49401, or to Edward Orehek, Department of cally meets the demands of the stressor.
Psychology, University of Groningen, Grote Kruisstraat 2/1, 9712 TS Although stress buffering theory dominates social support re-
Groningen, the Netherlands. E-mail: [email protected] or [email protected] search, the theory has important empirical limitations. The most
482
 RELATIONAL REGULATION THEORY 483

critical problem is that stress buffering is observed only inconsis- health. Thus, although this literature does not permit definitive
tently, compared to the easily replicable main effect between conclusions, coping and appraisal do not appear to explain per-
perceived support and mental health. For example, in Lakey and ceived support’s link to mental health at present.
Cronin’s (2008) comprehensive review of studies of social support Thus, we believe that stress buffering theory does not provide a
and major depressive disorder, nearly all studies found main ef- complete explanation for main effects between perceived support
fects, but only the seminal Brown and Harris (1978) study reported and mental health. We want to emphasize that we do not claim that
consistent evidence for stress buffering. For example, Wade and stress buffering does not occur. We merely claim that the well-
Kendler (2000) reported no evidence for stress buffering after replicated main effects between perceived support and mental
extensive testing with large samples. Beyond research on major health are not completely explained by stress and coping.
depression, many other studies have not observed stress buffering
effects (e.g., Beeble, Bybee, Sullivan, & Adams, 2009; Burton, Research and Theory on Main Effects
Stice, & Seeley, 2004; Procidano, 1992; Stroebe, Zech, Stroebe, &
Abakoumkin, 2005). Thus, many of the documented links between Main effects between perceived support and mental health are
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

perceived support and mental health reflect main effects rather highly replicable. Nearly all studies have found cross-sectional,
This document is copyrighted by the American Psychological Association or one of its allied publishers.

than stress buffering. main effects between low perceived support and major depression
Second, enacted support does not at present explain perceived (Lakey & Cronin, 2008). Likewise, meta-analyses have revealed
support’s link to mental health. One problem is that stress buffer- consistent main effects between low perceived support and post-
ing theory describes support perceptions as mostly veridical ac- traumatic stress disorder symptoms (Brewin et al., 2000) and
counts of specific supportive actions (Hobfoll, 2009; Lakey & nonspecific psychological distress (Finch et al., 1999; Procidano,
Cohen, 2000). Yet enacted support and perceived support are not 1992). Nearly all studies have found links between perceived
strongly correlated (Barrera, 1986; Dunkel-Schetter & Bennett, support and happiness (Lakey, in press).
1990; Goldsmith, 2004; Haber, Cohen, Lucas, & Baltes, 2007; Although well-established empirically, there has been compar-
Uchino, 2009). A second problem is that many studies have found atively little theoretical development to explain main effects. Co-
no link between enacted support and mental health (Barrera, 1986; hen and Wills (1985) sketched a few mechanisms. They suggested
Finch et al., 1999) or have found that receiving enacted support is that support might promote mental health by providing “persons
linked to worse mental health (Barrera, 1986; Bolger & Amarel, with regular positive experiences and a set of stable, socially
2007; Bolger, Zuckerman, & Kessler, 2000). One explanation for rewarded roles. . . . This kind of support . . . provides positive
these counterintuitive findings is that the link between enacted affect, a sense of predictability and stability . . . and . . . self-worth”
support and mental health is obscured because people with high (Cohen & Wills, 1985, p. 311). Thoits (1985) developed a sym-
stress receive the most enacted support and also have the worst bolic interactionist account of main effects that focused on the
mental health (Barrera, 1986; Larzelere, Kuhn, & Johnson, 2004). importance of identity and roles. Social roles (e.g., spouse, teacher)
Yet Seidman, Shrout, and Bolger’s (2006) simulations suggest that provide frameworks for social interaction that promote identity,
such an explanation requires implausibly strong links between belonging, self-esteem, and opportunities for mastery. According
stress and worse mental health. Thus, given that enacted support is to Thoits, “aspects of regularized social interaction and not emo-
not consistently linked to better mental health, it does not appear tional support dimensions per se, are responsible for maintaining
to be the mechanism linking perceived support to mental health. well-being. What we recognize as dimensions of emotional sup-
Third, coping and appraisal do not at present explain the link port and main effects of support are simply byproducts of these
between perceived support and mental health. Surprisingly, there more abstract social-psychological processes” (Thoits, 1985, pp.
is a lack of a focused, coherent literature on this question (Lakey 57–58).
& Cohen, 2000). If coping or appraisal accounted for perceived
support’s link to mental health, controlling for coping or appraisal Relational Regulation Theory
would substantially reduce this link (Baron & Kenny, 1986). Few
studies on coping have shown this pattern (see Holahan, Moos, Here, we describe relational regulation theory (RRT), developed
Holahan, & Brennan, 1995, for an exception). In contrast, many to explain the main effects between perceived support and mental
studies have found no evidence that coping can explain perceived health. RRT was developed inductively in the course of a 30-year
support’s link to mental health (Ben-Zur & Michael, 2007; Frazier, research program in which many of the effects predicted by stress
Tix, Klein, & Arikian, 2000; Giurgescu, Penckofer, Maurer, & buffering theory were consistently difficult to obtain. In contrast,
Bryant, 2006; Haley et al., 1996; Savelkoul, Post, de Witte, & van relational influences on perceived support and main effects be-
den Borne, 2000; Smith & Wallston, 1992). tween perceived support and mental health were strong and easily
The smaller literature on appraisal shows a similar pattern. With replicable. In addition to accommodating these two findings, RRT
few exceptions (Ben-Zur & Michael, 2007), appraisal has not been was developed to meet goals that we thought were valuable in any
able to explain perceived support’s link to mental health (Connell, social support theory. The theory should describe how perceived
Davis, Gallant, & Sharpe, 1994; Haley et al., 1996; Smith & support and mental health are rooted in social interaction and
Wallston, 1992). Moreover, even though theory specifically pre- should clearly distinguish between social and personality influ-
dicts stress buffering, this effect is rarely reported in the coping or ences.
appraisal literatures. Finally, coping theory and research have their RRT tackles the serious problem that standard methods do not
own empirical and theoretical challenges (Coyne & Gottlieb, 1996; unambiguously rule out the possibility that main effects reflect
Coyne & Racioppo, 2000) that present additional problems for recipient personality. Thus, interpretations of whether research
relying on coping to explain perceived support’s link to mental findings reflect personality or social processes primarily reflect the
 484 LAKEY AND OREHEK

theoretical preferences of observers. This problem is illustrated by ers are situations, relational influences are identical to Person ⫻
the following questionnaire item: “It helps to turn to my romantic Situation interactions in Endler and Hunt (1969) and essentially
partner in times of need.” Does this item reflect personality or similar to “if . . . then . . . , situation– behavior profiles” in Shoda,
social processes? A social support researcher might see it as a Mischel, and Wright (1994, p. 684) as well as Mischel and Shoda
perceived support item reflecting social processes. An attachment (1995, p. 250). RRT also adopts the SRM’s quantitative definitions
researcher might see it as an adult romantic attachment item of recipient and provider influences. Recipient influences are de-
reflecting personality. In fact, the item is from an attachment fined as mean differences among recipients, averaged across pro-
measure (Fraley, Waller, & Brennan, 2000). This ambiguity is a viders. Provider influences are defined as mean differences among
serious problem. Measuring and controlling for personality (e.g., providers averaged across recipients.
the Big Five) do not solve the problem because such measures The strengths of relational, recipient, and provider influences
reflect a blend of both social and personality influences (Paulhus & can be estimated from either generalizability (G) theory (Cron-
Reynolds, 1995). Thus, it is not possible to statistically control for bach, Gleser, Nanda, & Rajaratnam, 1972) or SRM designs
only personality variance using conventional measures of person- (Kenny, 1994; Kenny et al., 2006) using a variety of estimation
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

ality without also removing other sources of variance. procedures, such as ordinary least squares analysis of variance. As
This document is copyrighted by the American Psychological Association or one of its allied publishers.

applied to perceived support, G studies require that recipients each

Scope and Definitions rate a separate group of providers on supportiveness. In the SRM,
the typical study design is round robin, in which each member of
Scope. RRT was developed to explain perceived support’s a group rates each other. These two designs are depicted in Figure
link to emotional and affective disturbance in adults and adoles- 1. Although the definitions of recipient, provider, and relational
cents. The most relevant phenomena are described by diagnostic influences are essentially similar in G and SRM designs (Kenny et
criteria for anxiety and mood disorders, as well as self-report al., 2006), it is easier to describe these effects for the G design. A
measures of anxiety, depression, and general psychological dis- numerical example for the G design depicted in Figure 1 is
tress. These typically involve high negative affect and, often, low provided in Table 1.
positive affect (Watson, Clark, & Carey, 1988). RRT applies to In Table 1, Recipient A rates Provider B’s supportiveness as 9.
negative thoughts about the self, world, and future, given that Relational influences reflect the fact that Recipient A’s perception
depression and anxiety are strongly linked with such negative of Provider B’s supportiveness is 1.75 points higher than expected
thoughts (Beck, 1967; Dozois & Beck, 2008). RRT also applies to (i.e., 7.25) given that Recipient A’s typical rating of providers is 8
behavior associated with psychological distress (e.g., crying, sup- and Provider B’s typical rating by recipients is 6.5. Similarly,
port seeking, drug taking). Recipient B rates Provider B’s supportiveness as 4, which is 2.5
Quantitative definitions. We rely on Kenny and colleagues’ points lower than expected (i.e., 6.5) given that Recipient B’s
definition of relational from the social relations model (SRM; typical rating of providers is 6.5 and that Provider B’s typical
Kenny, 1994; Kenny, Kashy, & Cook, 2006). Relational influences rating by recipients is also 6.5.
occur when a provider elicits affect, action, or thought in a recip- The example just described has one observation per cell, and in
ient that is not characteristic of how the recipient typically re- such designs, the highest order interaction is confounded with
sponds to other providers and is not characteristic of what the sampling error (Kenny, 1994; Kenny et al., 2006). In this example,
provider typically elicits in other recipients.1 For example, Bob relational influences (i.e., the Recipient ⫻ Provider interaction) are
sees Kate as more supportive than how Bob typically sees other confounded with sampling error. Adding an additional factor (e.g.,
people and as more supportive than how others typically see Kate. items or time) permits sampling error to be shifted away from
As applied to affect, relational influences occur when, for ex- relational influences to the Recipients ⫻ Providers ⫻ Item (or
ample, a specific recipient experiences less sadness in response to
Time) effect, providing more accurate estimates of and signifi-
a specific provider than is typically experienced with other pro-
cance tests for relational influences (Kenny, 1994; Kenny et al.,
viders and less sadness than the provider typically elicits in other
2006). All of the studies on relational influences on perceived
recipients. With regard to action, a recipient drinks more than
support, reviewed momentarily, have separated relational influ-
usual when with a specific provider and more than the provider
ences from sampling error in this way.
typically elicits in other recipients. When perceived support and
The quantitative definition of relational influences has a key
affect are specifically relational, we refer to them as relational
advantage and intriguing implications. The key advantage is that
perceived support, relational affect, and so on.
relational influences are cleanly distinguished from recipient trait
When recipients’ reactions to the same providers are assessed
and provider influences. This solves the problem in social support
(i.e., each recipient is a level of a Recipients factor and each
research that the standard methods do not unambiguously indicate
provider is a level of a Providers factor), relational influences are
whether the link between perceived support and mental health
defined using the equations of the familiar general linear model
reflects relationships or personality.
(Kenny, 1994, p. 233; Kenny et al., 2006, pp. 194 –196). The
One intriguing implication of this quantitative definition is that
definition of relational influences (Rij) on a given variable for
it permits one to identify relational influences not only between
Recipient i and Provider j is Rij ⫽ Xij ⫺ Ri ⫺ Pj ⫹ M, where Xij
is Recipient i’s score on a variable with regard to Provider j, Ri is
Recipient i’s mean reaction across all providers, Pj is the mean 1
We have adopted the terms support recipient and support provider
reaction to Provider j across all recipients, and M is the grand because these terms seem to fit social support particularly well. Recipient
mean. Relational influences are conventional statistical interactions corresponds to actor or perceiver in the SRM, and provider corresponds to
for random effects (Kenny, 1994; Kenny et al., 2006). When provid- partner or target.
 RELATIONAL REGULATION THEORY 485

A B

Provider A Provider B Recipient A Recipient B

Provider A Provider B

Recipient A Recipient B
Recipient C Recipient D
Provider C Provider D
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Figure 1. Examples of generalizability and round-robin designs as used in social support research. A: A
This document is copyrighted by the American Psychological Association or one of its allied publishers.

generalizability design in which each recipient rates the supportiveness of each provider but providers do not rate
recipients. B: A round-robin design in which each recipient rates the others on supportiveness.

recipients and providers but also between recipients and activities, well as when the two reflect relational influences. However, it is
symbolic people (e.g., people presented only in photographs or on not possible to correlate relational support with trait depression
video), animals, ideas, and objects. To maintain the important because these influences are represented incommensurately. As
distinction between relational influences between people and re- indicated by the definitions of relational and recipient influences,
lational influences between people and nonpeople, we refer to the the entire Recipient ⫻ Provider matrix is required to represent
latter as quasi-relational influences or quasi relationships. Be- relational influences, whereas recipient influences are represented
cause relational influences are defined using equations that do not as a single column containing mean scores for recipients, averaged
require that providers are real people, it is meaningful to talk about across providers. As depicted in Table 1, it is not possible to
recipients’ quasi relationships with sports (i.e., activities), TV correlate a smaller number of scores arranged as a single column
characters (i.e., symbolic people), dogs (i.e., animals), music (i.e., (two in this example) with a larger number of scores arranged as
ideas), and cars (i.e., objects). As described subsequently, defining a Recipient ⫻ Provider matrix (four in this example).
relational and quasi-relational influences on the same metric per- When recipients rate the same providers, it is difficult to study
mits an understanding of how relational regulation is rooted in people’s most important personal relationships, as few recipients
similarities between the recipient’s and provider’s quasi relation- have the same important providers. Thus, in people’s natural
ships. This feature forms a key part of RRT’s description of how environments, providers are typically nested within recipients (an
perceived support is rooted in social interaction. example of a “one with many” design; Kenny et al., 2006). When
Defining relational influences quantitatively provides a way to providers are nested within recipients, relational and provider
think more contextually about constructs that are typically viewed influences are confounded in a single social influences component.
as traits. For example, by definition, major depressive disorder In such a design, social influences capture change in support and
(Diagnostic and Statistical Manual of Mental Disorders, 4th ed.; affect that result from interacting with or thinking about different
American Psychiatric Association, 1994) is a property of persons providers. Trait influences are defined as before: mean differences
that is stable across time and situations. However, thinking rela- among recipients in their reactions averaged across providers.
tionally requires one to consider how a person’s depressive symp- Relational regulation. RRT defines relational regulation as
toms ebb and flow depending upon with whom one interacts and desired affect, action, or thought that results from interacting with
in what activities one participates (cf. Lewinsohn & Graf, 1973). or thinking about specific other people. Quasi-relational regulation
RRT methods can estimate the links between perceived support occurs from interacting with or thinking about activities, symbolic
and depression for both the trait and relational aspects of both people, animals, ideas, or objects. Regulation is relational insofar
constructs. Thus, one can consider the link between low perceived as the provider elicits reactions in the recipient beyond how other
support and depression when the two reflect recipients’ traits as recipients typically respond to the provider and beyond how the

Table 1
A Numerical Example of Relational Influences

Deviation from Deviation from Recipient

Recipients/provider mean Provider A expectation Provider B expectation mean

Recipient A 7 ⫺1 9 1.75 8
Recipient B 9 1.75 4 ⫺2.5 6.5
Provider mean 8 6.5

Note. Boldface indicates recipient or provider means. Italics indicate the extent to which the observed score
differs from the score that would be expected given the corresponding recipient and provider mean.
 486 LAKEY AND OREHEK

recipient typically responds to other providers. We assume that (Mischel & Shoda, 1995; Shoda et al., 1994) did for situations and
people have preferred levels of affect, action, and thought for childhood behavior.
specific contexts and that people are motivated to achieve and In contrast, stress buffering theory is nomothetic in describing
maintain those preferred levels. What is desired depends on the objectively supportive actions for normatively defined stressful
person. A provider’s desired levels of talkativeness might conflict situations. For example, the optimal matching hypothesis of stress
with a recipient’s desired levels. buffering theory predicts that problem-solving support is most
Relational regulation can occur within any time frame, ranging effective for controllable stressors and emotional support is most
from seconds to hours to years. At one extreme, one might switch effective for uncontrollable stressors (Cohen & Hoberman, 1983;
conversation partners within seconds. At another extreme, one Cutrona & Russell, 1990).
might build a career over the course of decades. Relational regu- Yet both RRT and stress buffering theory are nomothetic at the
lation can be controlled or automatic (Gallo, Keil, McCulloch, level of the correlation between perceived support and mental
Rockstroh, & Gollwitzer, 2009; J. J. Gross, 1998). For example,
health. On average, people with higher perceived support will have
switching careers might require years of planning and preparation,
better mental health than people with lower perceived support.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

whereas channel surfing might involve little conscious awareness

There is evidence for strong relational influences on perceived
This document is copyrighted by the American Psychological Association or one of its allied publishers.

and resource consumption. Social interaction can be launched by a

support. The best evidence comes from studies in which recipients
person seeking to regulate him- or herself or by a person seeking
rated providers who were well known to recipients. Using G
to regulate another (e.g., when a parent regulates a child). Regu-
lation among people is typically reciprocal in that interaction designs, Lakey, McCabe, Fisicaro, and Drew (1996) asked doc-
launched by a recipient will also influence the affect, thought, and toral students to rate the supportiveness of program faculty (Study
actions of the provider, which will in turn influence the recipient. 1), and sorority members rated randomly selected sorority sisters
(Study 2). Giblin and Lakey (2010) asked medical fellows to rate
their clinical mentors. Branje, van Aken, and van Lieshout (2002),
Key Principles of Relational Regulation Theory as well as Lanz, Tagliabue, and Rosnati (2004), used round-robin
1. Recipients regulate their affect, action, and thought pri- designs. In Branje et al., four-person, Dutch, nuclear families rated
marily through social interaction. RRT accepts a key feature each family member. In Lanz et al., the families were Italian.
of attachment theory (Bowlby, 1969) and the need to belong Perceived support was overwhelmingly relational in each of these
hypothesis (Baumeister & Leary, 1995): People need a few ongo- studies.2
ing personal relationships to maintain their emotional well-being. Lakey’s (2010) error-corrected, meta-analytic estimate from
Thus, the link between perceived support and mental health pri- these five studies, which included over 5,000 dyads, indicated that
marily reflects social interaction, although thought alone about relational influences accounted for 62% of the systematic variance.
providers can influence mental health. RRT and stress buffering Recipient trait influences were also strong, accounting for 27%.
theory are similar in focusing on the importance of personal Provider influences were comparatively weak, accounting for only
relationships and social interaction for human well-being. How- 7%. In addition to studies of providers well known to recipients,
ever, stress buffering theory emphasizes enacted support, and as other studies have found large relational influences when symbolic
described later, RRT emphasizes ordinary yet affectively conse- providers were presented only via video (Cooper, Lakey, &
quential social interaction. Their shared focus on social interaction Cronin, 2010; Lakey, Cohen, & Neely, 2008; Lakey, Drew, & Sirl,
differs from theories that describe the traitlike aspect of perceived 1999; Lakey, McCabe, et al., 1996, Study 3) and when real
support (Lakey & Cassady, 1990; Sarason, Pierce, & Sarason, providers were strangers to recipients (Neely et al., 2006; Veenstra
1990; Uchino, 2009). Evidence for this principle is not unique to
et al., in press).
RRT and is provided by the voluminous research on the link
There is also evidence for relational influences on mental health.
between perceived support and mental health (Barrera, 1986;
Studies of this process have necessarily focused on short-term
Brewin et al., 2000; Cohen & Wills, 1985; Finch et al., 1999;
changes in recipients’ affect, given the practical and ethical diffi-
Lakey, in press; Lakey & Cronin, 2008).
culties involved in producing clinically significant changes in
2. Social interaction primarily regulates affect, action, and
thought relationally. Phrased differently, the specific provider mental health. Recipients have rated affect in response to providers
who successfully regulates a specific recipient is primarily a matter in clinical training and psychotherapy groups (Ingraham & Wright,
of personal taste. For example, one provider’s stoicism and another 1987) and in brief face-to-face conversations (Levesque & Kenny,
provider’s emotional expressiveness will be effective for different 1993; Neely et al., 2006; Veenstra et al., in press), to video
recipients. Regulation through interacting with activities, symbolic demonstrations of psychotherapists’ techniques (Lakey et al.,
people, animals, ideas, and objects is also largely relational. That 2008, Study 2), to symbolic providers presented via video inter-
is, people differ greatly in their affective, cognitive, and behavioral views (Cooper et al., 2010, Study 3), and to professors’ lectures (J.
reactions to different sports, TV characters, dogs, music, and cars.
By emphasizing relational influences, RRT is explicitly idio- 2
Both Branje et al. (2002) and Lanz et al. (2004) calculated relational
graphic at one level. That is, recipients’ profiles of reactions to the
influences separately for all 12 dyads derived from the combinations of the
same providers vary greatly across recipients in a way that is not roles of mothers, fathers, older children, and younger children. Provider
captured well by the objective characteristics of providers or their and recipient influences were also calculated separately for each type of
actions. Instead, investigators must map each recipient’s profile of family role. The values for relational, recipient, and provider influences
reactions across providers and other stimuli as Endler and Hunt reported here and in Lakey (2010) combined the effects across dyads for
(1969) did for situations and anxiety and Shoda and colleagues relational influences and across roles for recipient and provider influences.
 RELATIONAL REGULATION THEORY 487

Gross, Lakey, Edinger, Orehek, & Heffron, 2009). Each study Nonetheless, RRT hypothesizes that such regulation plays an
found significant relational influences on recipient affect. important role in stressful situations as well. For example, when
More ecologically valid studies have assessed affect elicited by awaiting news of the outcome of a relative’s surgery, people will
recipients’ important providers. As described previously, in such rely heavily upon the ordinary social interaction that helps regulate
designs, providers are nested within recipients because each recip- them in nonstressful situations (e.g., discussing family members’
ient has different important providers. Thus, relational and pro- exploits, work, or sports). Thus, in surgery waiting rooms, one
vider influences are combined into a single component that is should observe large amounts of ordinary talk as well as troubles
referred to here as social influences.3 In seven studies (Barry, talk.
Lakey, & Orehek, 2007; Lakey, Orehek, Hain, & VanVleet, 2010; Consistent with these predictions, among mothers awaiting their
Lakey & Scoboria, 2005), there were strong social influences on infants’ heart surgeries, generic relationship quality, but not en-
both positive and negative affect, accounting for nearly 60% of the acted support accounted for perceived support’s link to low dis-
error-corrected variance, with recipient trait variance accounting tress (Kaul & Lakey, 2003). Mak, Bond, Simpson, and Rholes
for the remaining 40%. Merlo and Lakey (2007) found the same (2010) replicated these findings among students. Hicks and Dia-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

pattern for subclinical depressive symptoms. mond (2008) observed that discussing positive events was linked
This document is copyrighted by the American Psychological Association or one of its allied publishers.

According to RRT, perceived support and mental health should to greater well-being but that discussing stressors was not. Clark,
be correlated when constructs reflect relational influences. Testing MacGeorge, and Robinson (2008) reported that the offer of com-
this hypothesis requires isolating relational, recipient, and provider panionship was preferred by participants to the types of support
influences and then estimating the correlations among constructs emphasized by stress buffering theory. Rook (1987) found that
for relational influences. Cronbach et al.’s (1972) multivariate G companionship was more closely linked to well-being than was
analyses are well suited for this purpose. In two studies, recipients enacted support. Hays and Oxley (1986) observed that interactions
interacted with the same providers over multiple occasions and involving fun and relaxation were linked to adjustment but that the
rated provider supportiveness as well as their own affect during the support emphasized by stress buffering theory was not. Capital-
conversation (Neely et al., 2006; Veenstra et al., in press). Partic- ization support in response to good events was linked to emotional
ipants were strangers when the studies began so as to mimic well-being (Gable et al., 2004, 2006), was substantially correlated
common social support interventions. In both studies, provider with perceived support, and could explain part of perceived sup-
supportiveness and recipient positive affect were strongly linked port’s link to mental health (Shorey & Lakey, in press).
when correlations reflected relational influences. There were no Mehl, Vazire, Holleran, and Clark (2010) recorded conversa-
relational influences on negative affect in these studies, likely tions among students as they went about their daily lives. Conver-
because providers were strangers to recipients. Other studies asked sations were classified as small talk or substantive. Small talk
participants to rate support and affect elicited by participants’ involved only trivial information, whereas substantive conversa-
important providers4 (Barry et al., 2007; Lakey et al., 2010; Lakey tion involved meaningful information. For example, two music
& Scoboria, 2005). Each of these studies found strong correlations students talking about whether free jazz is really jazz or two
between provider support and recipients’ positive and low negative psychology students talking about whether Freud’s thought is
affect for social influences (i.e., provider plus relational influ-
ences). 3
One limitation of the nested design is that the distinction between
RRT predicts that relational regulation applies to negative think-
recipient influences and social influences is less clear than in the fully
ing (Beck, 1967; Dozois & Beck, 2008). In a number of studies, crossed design in which each recipient rates the same providers. In the fully
recipients reported their negative thinking when with important crossed design, one can be confident that recipient influences reflect
providers. Strong social influences were observed for low self- individual differences because each recipient is exposed to the same
esteem (Barry et al., 2007; Lakey & Scoboria, 2005) and self- stimulus conditions (i.e., providers). However, in the nested design, recip-
discrepancies (Barry et al., 2007), as well as automatic negative ient influences might also reflect differences in providers. For example,
thoughts, dysfunctional attitudes, worry, hopelessness, and perfec- Doug might have lower average support scores across his important pro-
tionistic thinking (Lakey & Tanner, 2010). Moreover, the provid- viders than does Annmarie across her providers. This might result because
Doug’s providers are objectively less supportive than Annmarie’s. In this
ers who elicited negative thinking also elicited less favorable affect
case, social influences would be wrongly assigned to trait influences.
and lower perceived support. Fortunately, the extent to which variance is wrongly assigned can be tested.
3. Relational regulation occurs primarily in ordinary yet If the nested design wrongly assigns social influences to recipient influ-
affectively consequential social interaction. By ordinary, we ences, then recipient influences should be stronger in the nested design than
mean day-to-day interactions in contrast to stress buffering theo- in the fully crossed design. There are a number of fully crossed and nested
ry’s emphasis on discussions of stress and how to cope with it (i.e., studies of perceived support, and contrary to the wrongly assigned hypoth-
troubles talk; Goldsmith, 2004). Ordinary social interaction in- esis, recipient influences in nested designs are slightly smaller (21%;
cludes discussion of positive events, as described by Gable and Lakey, in press) than those in fully crossed designs (27%; Lakey, 2010).
Thus, the nested design does not appear to wrongly assign social influences
colleagues’ work on capitalization support (Gable, Gonzaga, &
to trait influences, at least for perceived support.
Strachman, 2006; Gable, Reis, Impett, & Asher, 2004), as well as 4
Although these studies combined relational and provider influences
aspects of life that would be considered routine (e.g., the non-
into a single social influence component, our current meta-analytic esti-
stressful events of the typical day). Thoits’s (1985) hypothesis that mate from five studies indicates that relational influences are approxi-
the link between social support and well-being reflects “aspects of mately nine times stronger than provider influences (Lakey, 2010), and
regularized social interaction and not emotional support dimen- thus, the bulk of social influences on perceived support likely reflects
sions per se” (p. 57) captures well this aspect of RRT. relational influences.
 488 LAKEY AND OREHEK

psychology or philosophy would both be classified as substantive whom I play jazz”) and provide a basis for later relational regu-
(M. R. Mehl, personal communication, March 10, 2010). Time lation in subsequent conversations.
spent in small talk was linked to less favorable affect, but time in Relational regulation is more effective when reciprocal elabo-
substantive conversations was linked to more favorable affect. ration occurs. Reciprocal elaboration begins when a provider re-
The studies just described are consistent with the hypothesis that sponds to a recipient’s statement with new information consistent
perceived support’s main effects with mental health are based with the recipient’s relationships and quasi relationships. Consider
primarily on ordinary social interaction. Yet it is difficult to the following example in which two people elaborate on their
completely eliminate the possibility that main effects do not some- quasi relationships with jazz. Richard might say, “I bought this
how reflect enacted support during troubles talk. Perhaps enacted Miles Davis–Charles Mingus record that’s really interesting.”
support has not been measured adequately or studied under the Richard says this because thinking about the record elicits favor-
right circumstances. Studies are needed that test the relational link able affect in him and he has the expectation that the statement will
between perceived support and favorable affect when enacted elicit a similar reaction in Stephen. Understanding Richard’s state-
support could not have occurred. RRT’s hypotheses about the role ment launches spreading activation of Stephen’s representations of
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

of symbolic providers (e.g., celebrities or public figures) in rela- jazz, and if Stephen’s quasi relationships are consistent with Rich-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

tional regulation are useful here because RRT predicts that per- ard’s, Stephen will also experience favorable affect. Yet, because
ceived support is linked to affect relationally, when recipients view Stephen’s representations are somewhat different from Richard’s,
or think about symbolic providers. Symbolic providers mimic how Stephen will provide elaboration: “I didn’t know they collaborated.
real providers regulate recipients’ affect when recipients observe I bet it’s great!” Stephen’s response leads to increased favorable
symbolic providers’ participating in conversations and shared ac- affect in Richard and additional spreading activation, and Richard
tivities with other symbolic providers. To our knowledge, stress provides elaboration: “Actually, it’s horrible. Miles and Mingus
buffering theory does not make predictions about symbolic pro- are not compatible at all. Even though they play mostly ballads,
viders. Stress buffering theory appears to imply that rating the you can hear the tension in every measure.” This elaboration leads
supportiveness of never-met, symbolic providers would not be to enhanced favorable affect in both participants and prompts
meaningful because never-met, symbolic providers do not provide further elaboration. Of course, if Stephen had responded to Rich-
enacted support to recipients.
ard’s initial statement with “Jazz is boring; I like the music of
Cooper et al. (2010, Study 1) compared ratings of important
Britney Spears,” the conversation’s ability to regulate affect would
network members to those of symbolic providers. Students rated
have taken a different turn. Thus, RRT predicts that the correlation
mothers, fathers, and closest peers on support and affect elicited. In
between perceived support and mental health is rooted in conver-
addition, recipients rated the symbolic providers to which partic-
sations that provide a social context for lingering with and elabo-
ipants had regular exposure. Results for real providers were es-
rating on representations of relationships and quasi relationships.
sentially similar to those reported in other studies (Barry et al.,
Thus, RRT integrates the mechanisms for regulating affect through
2007; Lakey et al., 2010; Lakey & Scoboria, 2005): Perceived
social interaction with those for regulating affect through activi-
support was strongly socially influenced, affect was approximately
ties, symbolic people, animals, ideas, and objects.
equally influenced by social factors as by traits, and there were
Following Heider’s (1958) balance theory, relational regulation
significant correlations between support and affect when correla-
tions reflected social influences. The results for symbolic providers is most effective if conversation partners elaborate on other people
were nearly identical. In Cooper et al.’s Studies 2 and 3, recipients they view similarly. Thus, one reason why people talk about other
each rated the same symbolic providers, and strong relational people is because such talk provides a social context for relational
influences emerged for expected supportiveness, which were regulation. Gossip about symbolic people (e.g., celebrities) should
linked to high positive and low negative affect. Thus, Cooper et al. work the same way. Thus, the popularity of supermarket tabloids
found the expected links between perceived support and affect for reflects a key mechanism by which many people regulate affect.
relational and social influences when enacted support and troubles As indicated by research on similarity and attraction (Byrne,
talk could not have occurred. 1971; Huston & Levinger, 1978), dyads with similar relationships
4. Relational regulation occurs primarily through conversa- and quasi relationships should be more effective in regulating each
tion and shared activities that elaborate on recipients’ cogni- other. Thus, similarities in attitudes, values, and activities should
tive representations of relationships and quasi relationships. be good markers of effective regulation, leading dyad members to
Each recipient has cognitive representations of relationships and perceive each other as supportive. Yet the link between similarity
quasi relationships (Andersen & Chen, 2002), and these represen- and relational regulation should be complex (cf. Montoya, Horton,
tations are linked to affect and action tendencies (Andersen & & Kirchner, 2008). For example, RRT predicts that relational
Chen, 2002; Bargh & Ferguson, 2000; Bower, 1981). Conversa- support is rooted not only in whether a recipient has similar
tion partners provide social contexts to support their experience of attitudes to a provider but primarily in the extent to which talking
favorable affect and thought by activating elements of their cog- about these attitudes elicits favorable affect. For example, two
nitive representations of relationships and quasi relationships. The people might both have similar religious beliefs, but talking about
provider helps to activate representations by listening, making their beliefs might not elicit favorable affect. If so, the two would
comments, and asking questions. In shared activities (e.g., playing not be expected to regulate each other by talking about religion. In
music), recipients and providers participate together in quasi rela- addition, two people might be dissimilar on many attitudes, but
tionships that elicit similarly favorable affect and thought in both talking about a handful might be highly effective. If so, the two
of them. Such affect and thought become linked to recipients’ and people will regulate each other well if they can stick to the
providers’ representations of each other (e.g., “Richard, with effective topics.
 RELATIONAL REGULATION THEORY 489

However, when social settings constrain reciprocal elaboration, tive when discussing upsetting events, the recipient sees the pro-
relational regulation will also be constrained. For example, two vider as even more supportive. The most effective relationships are
people might regulate each other well by talking about religion and those that regulate affect in the widest range of circumstances.
politics. Yet, if the social setting constrains these topics, they However, if the provider is not effective in regulating affect for
might never successfully regulate each other and might instead upsetting events, then the provider’s supportiveness is revised
regulate themselves with more beer and deep-fried, extra-salty lard downward. Of course, a provider who is initially judged as unsup-
sticks. Of course, religion and politics are commonly excluded portive is avoided, and thus, recipients do not disclose upsetting
from polite conversation because talking about these topics often events. As a result, the recipient’s initial judgment of unsupportive
dysregulates people. is unlikely to be revised. Ultimately, perceived support is inferred
Relational regulation might also occur during troubles talk. Recip- from affect regulation, and affect regulation is derived from social
ients have personal tastes about their preferred style of talking, think- interaction. In this sense, support judgments are secondary to
ing, feeling, and acting with regard to bad events. Troubles talk helps affect, a prediction that differs from appraisal models that predict
regulate recipients and providers when they are well matched in their that supportive appraisals lead to affect (Cohen & Wills, 1985).
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

tastes in responding to bad events. For example, some people like cool An interesting research question identified by RRT concerns
This document is copyrighted by the American Psychological Association or one of its allied publishers.

rationality, whereas others like hot expressiveness. As an example of what happens when a provider who has been effective in helping
these different styles, and how they are likely rooted in culture, one of regulate a recipient in ordinary social interaction provides ineffec-
us once attended two funerals in Detroit, Michigan, separated by only tive enacted support at a crucial point. Recipients likely vary in the
a few days and miles. One funeral was in a storefront Baptist church extent to which they will revise support judgments in response to
in which relatives of the deceased expressed their grief intensely and disappointing enacted support, given that recipients vary in the
elaborately. The other was in a stone, Anglican church in which grief extent to which they weigh other provider characteristics in mak-
was expressed in a reserved and measured style. According to RRT, ing support judgments (Lutz & Lakey, 2001). Some recipients
these two styles will help regulate some people but will dysregulate might view the disappointing provider as completely unsupportive.
others. Thus, part of what happens when support providers help Other recipients might change their support judgments minimally.
regulate recipients during troubles talk is that providers make avail- Other recipients might compartmentalize relationships whereby
able a social context for recipients to express their affect and thought some providers are useful for regulating affect in many contexts
in a style that is relationally familiar and comforting. but not for discussing bad events. Recipients might also view
Research on how relational regulation is rooted in social interaction providers as supportive for some bad events but not other events.
is less well developed than for other RRT principles, yet preliminary Thus, RRT predicts that recipients should make meaningful
evidence is consistent with predictions. Perceived similarities between support judgments about providers very early in an acquaintance,
recipients and providers in attitudes, values, and life experiences are before enacted support is offered. Moreover, the information used
among the strongest markers of provider supportiveness (Lakey et al., to judge support early in an acquaintance should be the same
2002; Lakey, Ross, Butler, & Bentley, 1996; Suitor, Pillemer, & information used when recipients have known providers for many
Keeton, 1995; Westmaas & Cohen Silver, 2006). This link occurs years. As just described, provider similarity strongly predicts per-
when similarity is experimentally manipulated (Lakey, Ross, et al., ceived supportiveness. This link occurs among long-standing dy-
1996, Study 2) as well as specifically for relational influences (Lakey ads (Lakey et al., 2002; Lakey, Ross, et al., 1996, Study 1),
et al., 2008, Study 1; Lakey, Lutz, & Scoboria, 2004; Neely et al., strangers (Lakey, Ross, et al., 1996, Study 3; Neely et al., 2006),
2006). Consistent with RRT, dyad members’ similarity in emotional when recipients judge symbolic providers (Lakey et al., 2004,
reactions to situations is linked to relationship satisfaction (Anderson, 2008, Study 1), and when recipients know only providers’ attitudes
Keltner, & John, 2003). and values (Lakey, Ross, et al., 1996, Study 2). Recipients also
More recently, Sain and Lakey (2011) developed a measure of judge support from perceived provider agreeableness, and consis-
conversations that elaborate on recipients’ relationships and quasi tent with RRT, this link occurs among long-standing dyads (Lakey
relationships. Items include “I enjoy talking with her/him because et al., 2002), when recipients rate symbolic providers (Lakey et al.,
we have interesting conversations that last a long time” and “It is 2004), and when providers are represented only by personality
difficult to find something he/she and I both want to talk about.” profiles (Lutz & Lakey, 2001).
Conversation elaboration was correlated strongly with both per- 6. Relational regulation is dynamic in that people shift con-
ceived support and favorable affect and could account for a large versations, interaction partners, and activities in an attempt to
percentage of the link between support and affect. optimally regulate affect. Surfing TV channels to find sym-
5. Perceived support is based primarily on relational regu- bolic people to regulate affect is a specific instance of this general
lation of affect through ordinary interactions but sometimes principle. In addition, the extent to which recipients and providers
also on enacted support. At initial acquaintance, a recipient regulate each other changes over time. This instability results from
infers a provider’s supportiveness from expectancies about changes in recipients’ and providers’ relationships and quasi rela-
whether the provider will help regulate the recipient’s affect. These tionships. As dyad members change with respect to with whom
expectancies are based on information about the provider’s simi- and what they regulate themselves, the ability of the dyad members
larity to the recipient in their relationships and quasi relationships. to regulate each other will change as well. In dyads lasting de-
Preliminary support judgments are revised as a result of direct cades, there should be many shifts in the effectiveness of relational
experience with the provider. If a provider regulates a recipient’s regulation.
affect well when discussing ordinary events, the recipient will In Neely et al. (2006) and Veenstra et al. (in press), recipients
disclose increasingly more personal thoughts and experiences, had multiple conversations with the same providers across months
including upsetting experiences. If affect regulation remains effec- and weeks. This design isolated relational influences that were
 490 LAKEY AND OREHEK

stable across conversations from those that changed dynamically influences were observed. Hoyt (2002) also found large relational
from conversation to conversation. Both studies found significant influences among students who viewed video therapy demonstra-
stable and dynamic relational influences on perceived support and tions. Lakey and Ondersma (2008) found large quasi-relational
affect. Dynamic relational influences were not mere randomness or influences in clients’ responses to different components of a mo-
error as perceived support and affect were significantly correlated tivational intervention to reduce substance abuse.
for this influence. If dynamic relational influences reflected ran- 8. The wider the diversity of potential relationships and
domness, the two constructs should not have been correlated. quasi relationships that are available to recipients, the greater
7. Social support interventions will be more effective if they the likelihood of effective regulation. Conversely, people
harness relational regulation. Social support interventions should be less effective in regulation if there is a restricted range
have been less effective than hoped (Helgeson & Gottlieb, 2000; of providers available. This can be seen at the level of large groups
Hogan, Linden, & Najarian, 2002; Lakey & Lutz, 1996). This as well. The more freedom a society allows for choosing relation-
might be because interventions typically have reflected the implicit ships and quasi relationships, the more effective the people of the
assumption that many people are objectively supportive. In the society should be in relational regulation. Cities and the Internet
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

typical intervention, new providers were made available to at-risk convey important advantages in relational regulation because they
This document is copyrighted by the American Psychological Association or one of its allied publishers.

recipients. Presumably, investigators chose providers whom the make available a wider range of providers than do sparsely pop-
investigators thought were supportive. Yet, as described previ- ulated, isolated areas. Yet there will always be recipients who
ously, supportiveness primarily reflects relational influences. RRT cannot effectively regulate themselves through currently accessi-
predicts that social support interventions will be more effective if ble relationships or quasi relationships. Rather than concluding
the interventions focus on relational influences. that such failures in regulation reflect trait deficits on the part of
Interventions designed to harness relational influences would be recipients, RRT encourages asking the question “Can this person
designed differently than previous interventions. Rather than pro- not be regulated, or has he/she not yet found the relationships and
vide access to objectively supportive providers, relational inter- quasi relationships that will achieve regulation?” Posing this ques-
ventions would match recipients with providers such that support- tion encourages continued search for relationships and quasi rela-
ive relationships emerged. This requires accurately forecasting tionships that might regulate the recipient. To our knowledge, the
supportive matches. Cronbach et al.’s (1972) multivariate G anal- RRT diversity principle has not yet been tested.
yses are useful tools for such forecasting. Trait-based prediction To summarize, RRT attempts to explain main effects between
(Wiggins, 1973) cannot be used to forecast relational influences perceived support and mental health by describing how people
because, as described previously, recipient and relational influ- regulate their affect through ordinary yet affectively consequential
ences are represented incommensurately. As depicted in Table 1, a conversation and shared activities. Perceived support typically
single recipient column representing trait influences cannot be does not cause affect directly but emerges from the types of social
mapped onto an entire Recipient ⫻ Provider matrix representing interaction that successfully regulate affect. Affect regulation via
relational influences. Instead, forecasting relational influences re- social interaction is primarily relational in that the people and activ-
quires predicting each recipient’s unique profile of responses ities that regulate affect are largely a matter of personal taste. RRT
across providers from predictor variables that are also represented predicts that social support and psychotherapeutic interventions will
as profiles across providers (Veenstra et al., in press). be more successful if designed to reflect relational influences. Pre-
In two studies, Veenstra et al. (in press) demonstrated that it is liminary evidence is consistent with these predictions.
possible to forecast relational support. In Study 1, recipients had
three conversations with the same providers across several weeks. Methodological Recommendations for
In Study 2, recipients had five conversations across several Social Support Research
months. In both studies, providers were strangers to recipients
initially. After each conversation, recipients rated providers on In addition to RRT’s prescription to separate recipient trait from
supportiveness and reported their own affect. The predicted crite- relational influences, RRT also makes recommendations for per-
rion was relational support averaged across the last two conversa- ceived support measurement and for alternative designs for testing
tions. Veenstra et al. successfully forecasted the specific providers whether support-related social interactions have causal influences.
who would be relationally supportive on the basis of recipients’ Most social support research asks respondents to rate their social
reactions to initial 10- or 20-min conversations. Thus, it is possible networks in general, rather than to rate specific providers. General
to forecast relational support, and multivariate G analyses are measures are limited, as investigators cannot distinguish among
useful tools for such forecasting. recipient, provider, and relational influences. Thus, a correlation
RRT also predicts that relational regulation occurs in psycho- between a general measure of perceived support and mental health
logical therapy. Recall that to isolate relational influences, recip- reflects some unknown mixture of recipient personality and social
ients must be exposed to the same providers. Because of the influences. Worse, if respondents make summary judgments of
practical and ethical nuances of having therapy recipients (pa- their social networks by using heuristics that average supportive-
tients) receive treatment from multiple providers (therapists) dur- ness across different providers (cf. items), then, according to
ing the same period of time, Lakey et al. (2008, Study 1) presented reliability theory (Cronbach et al., 1972), general measures of
therapy patients with videos of therapists demonstrating their perceived support compound recipient trait variance by averaging
techniques. Patients rated the expected supportiveness of the ther- out effects due to providers. Thus, general measures of perceived
apists and their expected therapeutic alliance. There were large support are likely composed of large amounts of trait influences,
relational influences on both. In Study 2, Lakey et al. presented a an undesirable situation if one wants to study social influences. A
new set of therapist videos to students. Again, large relational better approach, if the investigator cannot assess multiple provid-
 RELATIONAL REGULATION THEORY 491

ers for each respondent, is to ask respondents to rate a single, most Links to Other Social Support Theories
important support provider. Such a rating also confounds recipient
and social influences, but at least the magnitude of recipient We have already discussed the relation between RRT and stress
influences is not exaggerated by averaging across multiple provid- buffering theory. It is also important to briefly touch upon RRT’s
ers and the magnitude of the confound is reasonably well known. relation to two other theoretical approaches to social support.
RRT also suggests an alternative method for testing potential As mentioned previously, Thoits (1985) hypothesized that main
causal links between support-relevant social interaction and mental effects between social support and mental health primarily reflect
health. Many investigators believe that prospective designs are ordinary social interaction rather than stress and coping. According
strongest for testing hypotheses that low perceived support causes to this view, social roles provide a sense of identity and belonging,
maintain self-esteem, and provide opportunities for mastery (Cast
mental health problems. Such designs attempt to forecast from low
& Burke, 2002). Accordingly, research from this perspective has
Time 1 perceived support the worsening of mental health from
emphasized the number of social ties and roles, in contrast to
Time 1 to Time 2 and are viewed as especially rigorous because
RRT’s emphasis on perceived support. This is important because
perceived support (the hypothesized cause) occurs before the
although the number of social ties and roles has important links to
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

worsening of mental health (the hypothesized effect). Prospective

This document is copyrighted by the American Psychological Association or one of its allied publishers.

physical health (Uchino, 2004, 2009), ties and roles are not typi-
designs are appropriate for stress buffering theory as the theory cally strongly linked to mental health or perceived support (Bar-
predicts that perceived support acts as a stable resource that rera, 1986). Thus, RRT and Thoits’s theory appear to apply to
protects people from worsened mental health resulting from sub- different social support phenomena.
sequent life events. In contrast, prospective designs are not appro- Recently, several authors have developed perspectives on social
priate for testing hypotheses that social interaction leads to rela- support that draw from attachment theory (Collins, Guichard,
tively immediate changes in both mental health and perceived Ford, & Feeney, 2004; Feeney, 2004; Mikulincer & Shaver, 2009).
support (Lakey & Cronin, 2008). This is because in the typical RRT is similar to these in that RRT draws from attachment theory
prospective design, Time 2 mental health is assessed months or the premise that people need a few, long-lasting personal relation-
years after Time 1 perceived support. Relatively immediate influ- ships to maintain emotional well-being (Bowlby, 1969). Yet RRT
ences would not be detected months or years later, especially if concerns itself with relational influences, whereas these attach-
Time 1 mental health is controlled statistically. Controlling for ment theories of social support primarily emphasize personality.
Time 1 mental health removes any shared variance between Time
1 mental health and Time 1 perceived support. Yet this is exactly
the covariance that would reflect any immediate impact of social Concluding Comments
interaction on both perceived support and mental health. Before closing, we should address two potential concerns about
In the RRT approach, evidence for a causal role of social RRT. First, recipient, provider, and relational influences might be
interaction is provided by observing changes in mental health as inextricably confounded in nature, and thus, isolating these influ-
recipients interact with different providers. As examples of such a ences would not faithfully capture social support. Reis, Capobi-
repeated measures experimental design, Neely et al. (2006) and anco, and Tsai (2002) made a similar point regarding interdepen-
Veenstra et al. (in press) assigned recipients to interact with the dence theory:
same providers across multiple occasions. In these studies, inter-
actions with providers caused changes in recipients’ affect and An advantage of this approach is in not separating the “person” and
perceived support. These changes were relational in that recipients the “situation” into discrete factors, as most theories do, but rather in
responded differently to the same providers. Relational influences conceptualizing person effects in terms of the individual’s response to,
on perceived support and affect were strongly correlated, suggest- and selection among, the possibilities afforded by situations. (Reis et
al., 2002, p. 822)
ing that provider supportiveness could explain the features of
social interaction that caused affect.
Although this concern has its merits, attempting to reduce
Research designs appropriate for RRT are also useful for ad-
complex phenomena to simpler components has a long tradition
dressing the hypothesis that the link between low perceived sup-
in science (e.g., elements in chemistry, factor analytic research
port and poor mental health reflects traitlike differences among
in psychology). Ultimately, the utility of our approach will
recipients. Phrased differently, dispositionally unhappy people
depend on its ability to make new predictions about and account
might be unhappy about everything, including their support pro- for variation in support and mental health. As reviewed in the
viders. Research methods for RRT distinguish between trait and current article, substantial evidence suggests our approach is prom-
relational influences and estimate correlations separately for each. ising. A second concern is how to interpret the comparatively
RRT research consistently has found strong links between per- small magnitudes of provider influences. This might mean that
ceived support and favorable affect for recipient trait influences there is little in the way of objectively supportive providers
(Lakey, 2010; Lakey & Scoboria, 2005; Neely et al., 2006). Thus, (Lakey, 2010). An alternative interpretation is that such objec-
part of the link between perceived support and mental health tively supportive providers do exist but that they are sufficiently
reflects the phenomenon whereby dispositionally unhappy people uncommon in the population that their effects are diluted when
are dispositionally unhappy with their support providers. Yet part combined with those of ordinary providers. For example, recipi-
of the link also reflects social interaction. RRT shows how links ents might disagree about the supportiveness of Mark and Doug,
between mental health and perceived support can reflect each of but everyone might agree that Anne is supportive. This is an
these influences simultaneously. important research question. However, regardless of which expla-
 492 LAKEY AND OREHEK

nation better accounts for the magnitude of provider influences, adjustment to stress. Journal of Personality and Social Psychology, 79,
large relational influences still require explanation. 953–961. doi:10.1037/0022-3514.79.6.953
In summary, RRT was developed to explain the main effects Bower, G. H. (1981). Mood and memory. American Psychologist, 36,
between perceived social support and mental health. RRT hypoth- 129 –148. doi:10.1037/0003-066X.36.2.129
esizes that perceived support is primarily relational in that who is Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York,
NY: Basic Books.
supportive and what is supportive are largely a matter of personal
Branje, S. J. T., van Aken, M. A. G., & van Lieshout, C. F. M. (2002).
taste. The link between perceived support and mental health pri- Relational support in families with adolescents. Journal of Family Psy-
marily emerges in ordinary yet affectively consequential conver- chology, 16, 351–362. doi:10.1037/0893-3200.16.3.351
sations about people’s daily lives, as well as in shared activities, Brewin, C. R., Andrews, B., & Valentine, J. D. (2000). Meta-analysis of
rather than in conversations about how to cope with stress. RRT risk factors for posttraumatic stress disorder in trauma-exposed adults.
adopts a quantitative definition of relationships that permits their Journal of Consulting and Clinical Psychology, 68, 748 –766. doi:
clean separation from personality. The quantitative definition also 10.1037/0022-006X.68.5.748
permits analyses of recipients’ quasi relationships with activities, Brown, G. W., & Harris, T. (1978). Social origins of depression: A study
of psychiatric disorder in women. New York, NY: Free Press.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

symbolic providers, animals, ideas, and objects. This is important

This document is copyrighted by the American Psychological Association or one of its allied publishers.

because relational regulation among people is rooted in each Burton, E., Stice, E., & Seeley, J. R. (2004). A prospective test of the
person’s relationships and quasi relationships. RRT makes a num- stress-buffering model of depression in adolescent girls: No support
once again. Journal of Consulting and Clinical Psychology, 72, 689 –
ber of new predictions about social support research and suggests
697. doi:10.1037/0022-006X.72.4.689
new approaches to social support interventions. Preliminary evi-
Byrne, D. (1971). The attraction paradigm. New York, NY: Academic
dence suggests that RRT fits the data sufficiently well to deserve Press.
additional theoretical and empirical development. Cast, A. D., & Burke, P. J. (2002). A theory of self-esteem. Social Forces,
80, 1041–1068. doi:10.1353/sof.2002.0003
Clark, R. A., MacGeorge, E. L., & Robinson, L. (2008). Evaluation of peer
References comforting strategies by children and adolescents. Human Communica-
tion Research, 34, 319 –345. doi:10.1111/j.1468-2958.2008.00323.x
American Psychiatric Association. (1994). Diagnostic and statistical man-
Cohen, S., & Hoberman, H. M. (1983). Positive events and social supports
ual of mental disorders (4th ed.). Washington, DC: Author.
as buffers of life change stress. Journal of Applied Social Psychology,
Andersen, S. M., & Chen, S. (2002). The relational self: An interpersonal
13, 99 –125. doi:10.1111/j.1559-1816.1983.tb02325.x
social– cognitive theory. Psychological Review, 109, 619 – 645. doi:
Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering
10.1037/0033-295X.109.4.619
hypothesis. Psychological Bulletin, 98, 310 –357. doi:10.1037/0033-
Anderson, C., Keltner, D., & John, O. P. (2003). Emotional convergence
2909.98.2.310
between people over time. Journal of Personality and Social Psychol-
Collins, N. L., Guichard, A. C., Ford, M. B., & Feeney, B. C. (2004).
ogy, 84, 1054 –1068. doi:10.1037/0022-3514.84.5.1054
Working models of attachment: New developments and emerging
Bargh, J. A., & Ferguson, M. L. (2000). Beyond behaviorism: On the
automaticity of higher mental processes. Psychological Bulletin, 126, themes. In S. W. Rholes & J. A. Simpson (Eds.), Adult attachment:
925–945. doi:10.1037/0033-2909.126.6.925 Theory, research, and clinical implications (pp. 196 –239). New York,
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable NY: Guilford Press.
distinction in social psychological research: Conceptual, strategic, and Connell, C. M., Davis, W. K., Gallant, M. P., & Sharpe, P. A. (1994).
statistical considerations. Journal of Personality and Social Psychology, Impact of social support, social cognitive variables, and perceived threat
51, 1173–1182. doi:10.1037/0022-3514.51.6.1173 on depression among adults with diabetes. Health Psychology, 13,
Barrera, M., Jr. (1986). Distinctions between social support concepts, 263–273. doi:10.1037/0278-6133.13.3.263
measures and models. American Journal of Community Psychology, 14, Cooper, C., Lakey, B., & Cronin, A. (2010). Symbolic providers regulate
413– 445. doi:10.1007/BF00922627 recipients relationally: Implications for perceived support’s link to
Barry, R., Lakey, B., & Orehek, E. (2007). Links among attachment mental health. Manuscript in preparation.
dimensions, affect, the self, and perceived support for broadly general- Coyne, J. C., & Gottlieb, B. H. (1996). The mismeasure of coping by
ized attachment styles and specific bonds. Personality and Social Psy- checklist. Journal of Personality, 64, 959 –991.
chology Bulletin, 33, 340 –353. doi:10.1177/0146167206296102 Coyne, J. C., & Racioppo, M. W. (2000). Never the twain shall meet?
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for Closing the gap between coping research and clinical intervention re-
interpersonal attachments as a fundamental human motivation. Psycho- search. American Psychologist, 55, 655– 664. doi:10.1037/0003-
logical Bulletin, 117, 497–529. doi:10.1037/0033-2909.117.3.497 066X.55.6.655
Beck, A. T. (1967). Depression: Causes and treatment. Philadelphia: Cronbach, L. J., Gleser, G. C., Nanda, H., & Rajaratnam, N. (1972). The
University of Pennsylvania Press. dependability of behavioral measurements: Theory of generalizability
Beeble, M. L., Bybee, D., Sullivan, C. M., & Adams, A. E. (2009). Main, for scores and profiles. New York, NY: Wiley.
mediating, and moderating effects of social support on the well-being of Cutrona, C. E., & Russell, D. W. (1990). Type of social support and
survivors of intimate partner violence across 2 years. Journal of Con- specific stress: Toward a theory of optimal matching. In B. R. Sarason,
sulting and Clinical Psychology, 77, 718 –729. doi:10.1037/a0016140 I. G. Sarason, & G. R. Pierce (Eds.), Social support: An interactional
Ben-Zur, H., & Michael, K. (2007). Burnout, social support, and coping at view (pp. 319 –366). New York, NY: Wiley.
work among social workers, psychologists and nurses: The role of Dozois, D. J. A., & Beck, A. T. (2008). Cognitive schemas, beliefs and
challenge/control appraisals. Social Work in Health Care, 45, 63– 82. assumptions. In K. S. Dobson & D. Dozois (Eds.), Risk factors for
doi:10.1300/J010v45n04_04 depression (pp. 121–143). San Diego, CA: Academic Press.
Bolger, N., & Amarel, D. (2007). Effects of social support visibility on Dunkel-Schetter, C., & Bennett, T. L. (1990). Differentiating the cognitive
adjustment to stress: Experimental evidence. Journal of Personality and and behavioral aspects of social support. In B. R. Sarason, I. G. Sarason,
Social Psychology, 92, 458 – 475. doi:10.1037/0022-3514.92.3.458 & G. R. Pierce (Eds.), Social support: An interactional view (pp.
Bolger, N., Zuckerman, A., & Kessler, R. C. (2000). Invisible support and 267–296). New York, NY: Wiley.
 RELATIONAL REGULATION THEORY 493

Endler, N. S., & Hunt, J. M. (1969). Generalizability of contributions from Hicks, A. M., & Diamond, L. M. (2008). How was your day? Couples’
sources of variance in the S-R Inventories of Anxiousness. Journal of affect when telling and hearing daily events. Personal Relationships, 15,
Personality, 37, 1–24. doi:10.1111/j.1467-6494.1969.tb01728.x 205–228. doi:10.1111/j.1475-6811.2008.00194.x
Feeney, B. C. (2004). A secure base: Responsive support of goal strivings Hobfoll, S. E. (2009). Social support: The movie. Journal of Social and
and exploration in adult intimate relationships. Journal of Personality Personal Relationships, 26, 93–101. doi:10.1177/0265407509105524
and Social Psychology, 87, 631– 648. doi:10.1037/0022-3514.87.5.631 Hogan, B. E., Linden, W., & Najarian, B. (2002). Social support interven-
Finch, J. F., Okun, M. A., Pool, G. J., & Ruehlman, L. S. (1999). A tions: Do they work? Clinical Psychology Review, 22, 381– 440. doi:
comparison of the influence of conflictual and supportive social inter- 10.1016/S0272-7358(01)00102-7
actions on psychological distress. Journal of Personality, 67, 581– 621. Holahan, C. J., Moos, R. H., Holahan, C. K., & Brennan, P. L. (1995).
doi:10.1111/1467-6494.00066 Social support, coping and depressive symptoms in a late-middle-aged
Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response sample of patients reporting cardiac illness. Health Psychology, 14,
theory analysis of self-report measures of adult attachment. Journal of 152–163. doi:10.1037/0278-6133.14.2.152
Personality and Social Psychology, 78, 350 –365. doi:10.1037/0022- Hoyt, W. T. (2002). Bias in participant ratings of psychotherapy process:
3514.78.2.350 An initial generalizability study. Journal of Counseling Psychology, 49,
Frazier, P. A., Tix, A. P., Klein, C. D., & Arikian, N. (2000). Testing 35– 46. doi:10.1037/0022-0167.49.1.35
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

theoretical models of the relations between social support, coping and Huston, T. L., & Levinger, G. (1978). Interpersonal attraction and rela-
adjustment to stressful life events. Journal of Social and Clinical Psy- tionships. Annual Review of Psychology, 29, 115–156. doi:10.1146/
chology, 19, 314 –335. doi:10.1521/jscp.2000.19.3.314 annurev.ps.29.020178.000555
Gable, S. L., Gonzaga, G. C., & Strachman, A. (2006). Will you be there Ingraham, L. J., & Wright, T. L. (1987). A social relations model test of
for me when things go right? Supportive responses to positive event Sullivan’s anxiety hypothesis. Journal of Personality and Social Psy-
disclosures. Journal of Personality and Social Psychology, 91, 904 –917. chology, 52, 1212–1218. doi:10.1037/0022-3514.52.6.1212
doi:10.1037/0022-3514.91.5.904 Kaul, M., & Lakey, B. (2003). Where is the support in perceived support?
Gable, S. L., Reis, H. T., Impett, E. A., & Asher, E. R. (2004). What do you The role of generic relationship satisfaction and enacted support in
do when things go right? The intrapersonal and interpersonal benefits of perceived support’s relation to low distress. Journal of Social and
sharing positive events. Journal of Personality and Social Psychology, Clinical Psychology, 22, 59 –78. doi:10.1521/jscp.22.1.59.22761
87, 228 –245. doi:10.1037/0022-3514.87.2.228 Kenny, D. A. (1994). Interpersonal perception: A social relations analysis.
Gallo, I. S., Keil, A., McCulloch, K. C., Rockstroh, B., & Gollwitzer, P. M. New York, NY: Guilford Press.
(2009). Strategic automation of emotion regulation. Journal of Person- Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Analysis of dyadic
ality and Social Psychology, 96, 11–31. doi:10.1037/a0013460 data. New York, NY: Guilford Press.
Giblin, F., & Lakey, B. (2010). Integrating mentoring and social support Lakey, B. (2010). Social– clinical approaches to social support suggest new
research within the context of stressful medical training. Journal of strategies for intervention. In J. E. Maddux & J. P. Tangney (Eds.),
Social and Clinical Psychology, 29, 771–796. doi:10.1521/ Social psychological foundations of clinical psychology (pp. 177–194).
jscp.2010.29.7.771 New York, NY: Guilford Press.
Giurgescu, C., Penckofer, S., Maurer, M. C., & Bryant, F. B. (2006). Lakey, B. (in press). Personality and relational processes in perceived
Impact of uncertainty, social support and prenatal coping on the psy- support and happiness. In I. Boniwell & S. David (Eds.), Oxford hand-
chological well-being of high-risk pregnant women. Nursing Research, book of happiness. Oxford, England: Oxford University Press.
55, 356 –365. doi:10.1097/00006199-200609000-00008 Lakey, B., Adams, K., Neely, L. Rhodes, G., Lutz, C. J., & Sielky, K.
Goldsmith, D. J. (2004). Communicating social support. New York, NY: (2002). Perceived support and low emotional distress: The role of
Cambridge University Press. doi:10.1017/CBO9780511606984 enacted support, dyad similarity and provider personality. Personality
Gross, J., Lakey, B., Edinger, K., Orehek, E., & Heffron, D. (2009). Person and Social Psychology Bulletin, 28, 1546 –1555. doi:10.1177/
perception in the college classroom: Accounting for tastes in students’ 014616702237582
evaluations of teaching effectiveness. Journal of Applied Social Psy- Lakey, B., & Cassady, P. B. (1990). Cognitive processes in perceived
chology, 39, 1609 –1638. doi:10.1111/j.1559-1816.2009.00497.x social support. Journal of Personality and Social Psychology, 59, 337–
Gross, J. J. (1998). The emerging field of emotion regulation: An integra- 343. doi:10.1037/0022-3514.59.2.337
tive review. Review of General Psychology, 2, 271–299. doi:10.1037/ Lakey, B., Cohen, J. L., & Neely, L. C. (2008). Perceived support and
1089-2680.2.3.271 relational influences on psychotherapy process constructs. Journal of
Haber, M. G., Cohen, J. L., Lucas, T., & Baltes, B. B. (2007). The Counseling Psychology, 55, 209 –220. doi:10.1037/0022-0167.55.2.209
relationship between self-reported received and perceived social sup- Lakey, B., & Cohen, S. (2000). Social support theory and selecting mea-
port: A meta-analytic review. American Journal of Community Psychol- sures of social support. In S. Cohen, L. U. Gordon, & B. H. Gottlieb
ogy, 39, 133–144. doi:10.1007/s10464-007-9100-9 (Eds.), Social support measurement and interventions: A guide for
Haley, W. E., Roth, D. L., Coleton, M. I., Ford, G. R., West, C. A. C., health and social scientists (pp. 29 –52). New York, NY: Oxford Uni-
Collins, R. P., & Isobe, T. L. (1996). Appraisal, coping, and social versity Press.
support as mediators of well-being in Black and White family caregivers Lakey, B., & Cronin, A. (2008). Low social support and major depression:
of patients with Alzheimer’s disease. Journal of Consulting and Clinical Research, theory and methodological issues. In K. S. Dobson & D.
Psychology, 64, 121–129. doi:10.1037/0022-006X.64.1.121 Dozois (Eds.), Risk factors for depression (pp. 385– 408). San Diego,
Hays, R. B., & Oxley, D. (1986). Social network development and func- CA: Academic Press. doi:10.1016/B978-0-08-045078-0.00017-4
tioning during a life transition. Journal of Personality and Social Psy- Lakey, B., Drew, J. B., & Sirl, K. (1999). Clinical depression and percep-
chology, 50, 305–313. doi:10.1037/0022-3514.50.2.305 tions of supportive others: A generalizability analysis. Cognitive Ther-
Heider, F. (1958). The psychology of interpersonal relations. New York, apy and Research, 23, 511–533. doi:10.1023/A:1018772421589
NY: Wiley. doi:10.1037/10628-000 Lakey, B., & Lutz, C. J. (1996). Social support and preventive and
Helgeson, V. S., & Gottlieb, B. H. (2000). Support groups. In S. Cohen, therapeutic interventions. In G. R. Pierce, B. R. Sarason, & I. G. Sarason
L. G. Underwood, & B. H. Gottlieb (Eds.), Social support measurement (Eds.), Handbook of social support and the family (pp. 435– 465). New
and intervention: A guide for health and social scientists (pp. 221–245). York, NY: Plenum Press.
New York, NY: Oxford University Press. Lakey, B., Lutz, C. J., & Scoboria, A. (2004). The information used to
 494 LAKEY AND OREHEK

judge supportiveness depends on whether the judgment reflects the invariance in personality structure. Psychological Review, 102, 246 –
personality of perceivers, the objective characteristics of targets, or their 268. doi:10.1037/0033-295X.102.2.246
unique relationship. Journal of Social and Clinical Psychology, 23, Montoya, R. M., Horton, R. S., & Kirchner, J. (2008). Is actual similarity
817– 835. doi:10.1521/jscp.23.6.817.54806 necessary for attraction? A meta-analysis of actual and perceived simi-
Lakey, B., McCabe, K., Fisicaro, S., & Drew, J. (1996). Personal and larity. Journal of Social and Personal Relationships, 25, 889 –922.
environmental determinants of social support: Three generalizability doi:10.1177/0265407508096700
studies. Journal of Personality and Social Psychology, 70, 1270 –1280. Neely, L. C., Lakey, B., Cohen, J. L., Barry, R., Orehek, E., Abeare, C. A.,
doi:10.1037/0022-3514.70.6.1270 & Mayer, W. (2006). Trait and social processes in the link between
Lakey, B., & Ondersma, S. (2008). A new approach for detecting patient– social support and affect: An experimental laboratory investigation.
treatment matching in psychological therapy. Journal of Social and Journal of Personality, 74, 1015–1046. doi:10.1111/j.1467-
Clinical Psychology, 27, 56 – 69. doi:10.1521/jscp.2008.27.1.56 6494.2006.00401.x
Lakey, B., Orehek, E., Hain, K., & VanVleet, M. (2010). Enacted support’s Paulhus, D. L., & Reynolds, S. (1995). Enhancing target variance in
links to negative affect and perceived support are more consistent with personality impressions: Highlighting the person in person perception.
theory when social influences are isolated from trait influences. Person- Journal of Personality and Social Psychology, 69, 1233–1242. doi:
ality and Social Psychology Bulletin, 36, 132–142. doi:10.1177/ 10.1037/0022-3514.69.6.1233
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

0146167209349375 Procidano, M. E. (1992). The nature of perceived social support: Findings

Lakey, B., Ross, L., Butler, C., & Bentley, K. (1996). Making social of meta-analytic studies. In C. D. Spielberger & J. N. Butcher (Eds.),
support judgments: The role of perceived similarity and conscientious- Advances in personality assessment (Vol. 9, pp. 1–26). Hillsdale, NJ:
ness. Journal of Social and Clinical Psychology, 15, 283–304. doi: Erlbaum.
10.1521/jscp.1996.15.3.283 Reis, H. T., Capobianco, A., & Tsai, F. (2002). Finding the person in
Lakey, B., & Scoboria, A. (2005). Trait and social influences in the links personal relationships. Journal of Personality, 70, 813– 850. doi:
among perceived social support, affect and self-esteem. Journal of 10.1111/1467-6494.05025
Personality, 73, 361–388. doi:10.1111/j.1467-6494.2005.00312.x Rook, K. S. (1987). Social support versus companionship: Effects on life
Lakey, B., & Tanner, S. (2010). The social context of negative thinking: stress, loneliness, and evaluation of others. Journal of Personality and
Social and trait influences on negative thinking, social support, and Social Psychology, 52, 1132–1147. doi:10.1037/0022-3514.52.6.1132
affect. Manuscript in preparation. Sain, T., & Lakey, B. (2011). [A measure of elaboration in conversation].
Lanz, M., Tagliabue, S., & Rosnati, R. (2004). Il social relations model Unpublished raw data, Grand Valley State University.
nello studio delle relazioni familiari [Social relations model in family Sarason, B. R., Pierce, G. R., & Sarason, I. G. (1990). Social support: The
relationships’ study]. Testing Psicometria Metodologia, 11, 197–214. sense of acceptance and the role of relationships. In B. R. Sarason, I. G.
Larzelere, R. E., Kuhn, B. R., & Johnson, B. (2004). Intervention selection Sarason, & G. R. Pierce (Eds.), Social support: An interactional view
bias: An underrecognized confound in intervention research. Psycholog- (pp. 97–128). New York, NY: Wiley.
ical Bulletin, 130, 289 –303. doi:10.1037/0033-2909.130.2.289 Savelkoul, M., Post, M. W. M., de Witte, L. P., & van den Borne, H. B.
Lazarus, R. S. (1966). Psychological stress and the coping process. New (2000). Social support, coping and subjective well-being in patients with
York, NY: McGraw-Hill. rheumatic diseases. Patient Education and Counseling, 39, 205–218.
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal and coping. New doi:10.1016/S0738-3991(99)00033-6
York, NY: Springer. Seidman, G., Shrout, P. E., & Bolger, N. (2006). Why is enacted support
Levesque, M. J., & Kenny, D. A. (1993). Accuracy of behavioral predic- associated with increased distress? Using simulation to test two possible
tions at zero acquaintance: A social relations analysis. Journal of Per- sources of spuriousness. Personality and Social Psychology Bulletin, 32,
sonality and Social Psychology, 65, 1178 –1187. doi:10.1037/0022- 52– 65. doi:10.1177/0146167205279582
3514.65.6.1178 Shoda, Y., Mischel, W., & Wright, J. C. (1994). Intraindividual stability in
Lewinsohn, P. M., & Graf, M. (1973). Pleasant activities and depression. the organization and patterning of behavior: Incorporating psychological
Journal of Consulting and Clinical Psychology, 41, 261–268. doi: situations into the idiographic analysis of personality. Journal of Per-
10.1037/h0035142 sonality and Social Psychology, 67, 674 – 687. doi:10.1037/0022-
Lutz, C. J., & Lakey, B. (2001). How people make support judgments: 3514.67.4.674
Individual differences in the traits used to infer supportiveness in others. Shorey, R. C., & Lakey, B. (in press). Perceived and capitalization support
Journal of Personality and Social Psychology, 81, 1070 –1079. doi: are substantially similar: Implications for social support theory. Person-
10.1037/0022-3514.81.6.1070 ality and Social Psychology Bulletin.
Mak, M. C. K., Bond, M. H., Simpson, J. A., & Rholes, W. S. (2010). Smith, C. A., & Wallston, K. A. (1992). Adaptation in patients with
Adult attachment, perceived support, and depressive symptoms in Chi- chronic rheumatoid arthritis: Application of a general model. Health
nese and American cultures. Journal of Social and Clinical Psychology, Psychology, 11, 151–162. doi:10.1037/0278-6133.11.3.151
29, 144 –165. doi:10.1521/jscp.2010.29.2.144 Stroebe, W., Zech, E., Stroebe, M. S. & Abakoumkin, G. (2005). Does
Mehl, M. R., Vazire, S., Holleran, S. E., & Clark, C. S. (2010). Eaves- social support help in bereavement? Journal of Social and Clinical
dropping on happiness: Well-being is related to having less small talk Psychology, 24, 1030 –1050. doi:10.1521/jscp.2005.24.7.1030
and more substantive conversations. Psychological Science, 21, 539 – Suitor, J. J., Pillemer, K., & Keeton, S. (1995). When experience counts:
541. doi:10.1177/0956797610362675 The effects of experiential and structural similarity on patterns of sup-
Merlo, L. J., & Lakey, B. (2007). Trait and social influences in the links port and interpersonal stress. Social Forces, 73, 1573–1588. doi:
among adolescent attachment, coping, and depressive symptoms. Jour- 10.2307/2580459
nal of Clinical Child and Adolescent Psychology, 36, 195–206. doi: Thoits, P. A. (1985). Social support and psychological well-being: Theo-
10.1080/15374410701277846 retical possibilities. In I. G. Sarason & B. R. Sarason (Eds.), Social
Mikulincer, M., & Shaver, P. R. (2009). An attachment and behavioral support: Theory, research and application (pp. 51–72). Dordrecht, the
systems perspective on social support. Journal of Social and Personal Netherlands: Martinus Nijhoff.
Relationships, 26, 7–19. doi:10.1177/0265407509105518 Thoits, P. A. (1986). Social support as coping assistance. Journal of
Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of Consulting and Clinical Psychology, 54, 416 – 423. doi:10.1037/0022-
personality: Reconceptualizing situations, dispositions, dynamics, and 006X.54.4.416
 RELATIONAL REGULATION THEORY 495

Uchino, B. N. (2004). Social support and physical health: Understanding Watson, D., Clark, L. A., & Carey, G. (1988). Positive and negative
the health consequences of our relationships. New Haven, CT: Yale affectivity and their relation to anxiety and depressive disorders. Journal
University Press. of Abnormal Psychology, 97, 346 –353. doi:10.1037/0021-
Uchino, B. N. (2009). Understanding the links between social support and 843X.97.3.346
physical health: A life-span perspective with emphasis on the separabil- Westmaas, J. L., & Cohen Silver, R. (2006). The role of perceived
ity of perceived and received support. Perspectives on Psychological similarity in supportive responses to victims of negative life events.
Science, 4, 236 –255. doi:10.1111/j.1745-6924.2009.01122.x Personality and Social Psychology Bulletin, 32, 1537–1546. doi:
Veenstra, A., Lakey, B., Cohen, J. L., Neely, L. C., Orehek, E., Barry, R. 10.1177/0146167206291874
& Abeare, C. (in press). Forecasting the providers who recipients will Wiggins, J. (1973). Personality and prediction: Principles of personality
see as unusually supportive. Personal Relationships. assessment. Reading, MA. Addison-Wesley.
Wade, T. D., & Kendler, K. S. (2000). Absence of interactions between
social support and stressful life events in the prediction of major de- Received June 28, 2010
pression and depressive symptomatology in women. Psychological Revision received February 24, 2011
Medicine, 30, 965–974. doi:10.1017/S0033291799002251 Accepted March 4, 2011 䡲
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

ORDER FORM Check enclosed (make payable to APA)

Start my 2011 subscription to Psychological Review Charge my: Visa MasterCard American Express
ISSN:0033-295X
Cardholder Name
___ $79.00 APA MEMBER / AFFILIATE Card No. Exp. Date
___ $179.00 INDIVIDUALNONMEMBER
Signature (Required for Charge)
___ $629.00 INSTITUTION
In DC and MD add 6% sales tax Billing Address
TOTALAMOUNT DUE $ Street
City State Zip
Subscription orders must be prepaid. Subscriptions are on a calendar
year basis only. Allow 4-6 weeks for delivery of the first issue. Call for international Daytime Phone
subscription rates. E-mail
SENDTHIS ORDER FORM TO
American Psychological Association Mail To
Subscriptions Name
750 First Street, NE
Washington, DC 20002-4242 Address

Call 800-374-2721 or 202-336-5600

Fax 202-336-5568 : TDD/TTY 202-336-6123 City State Zip
For subscription information, APA Member #
e-mail: [email protected]
REVA11